Summary of the Court's Decision:

By motion filed with the Court on August 4, 2025, the defendant moves for the Court to preclude expert witness testimony on the People's direct case regarding STRmix results, arguing that 1) the probative value of the testimony would be substantially outweighed by its prejudicial effect, 2) the results are unreliable and not accepted by the scientific community where the mixture involves potentially five (5) or more DNA contributors, and/or 3) the prosecution cannot lay a proper foundation.

In the alternative, he moves for a Frye, Wesley, Brooks hearing to determine the reliability of the proposed testimony and the underlying procedures used and/or the probative vs prejudicial effect. He also moves for the Court to give jury instructions on 1) the New York City Office of Chief Medical Examiner's ("OCME") own reported error rates for a four-person mixture and a three-person mixture, 2) expert witnesses lack of knowledge on how or when any DNA was deposited, and 3) that the jury may infer that the DNA of any possible contributor, including any possible DNA from the defendant, was deposited in a manner consistent with innocence, or was deposited at a different time and place than the time and place alleged.

The People oppose the defendant's motion to preclude expert testimony or for a Frye hearing regarding STRmix and OCME's determination of the estimated number of contributors. The People argue that "general acceptance" under Frye is demonstrated by the overwhelming number of laboratories which have adopted, or are in the processing of adopting, STRmix, and the United States and New York State Courts that have ratified the use of STRmix as meeting the test of general acceptance and reliability. The People further argue that the defendant overlooked the cases that conducted Frye or Daubert hearings and affirmed the general acceptance, reliability and admissibility of STRmix. The People aver that the defendant's arguments regarding the estimated number of contributors, goes to the weight of the evidence, not the reliability of the scientific principles, and that defendant can raise any such challenges [*2]during cross-examination of their expert and/or by testimony of his own expert.

On May 17, 2023, at approximately 2:45 AM, at the intersection of Merrick Boulevard and 116th Avenue, Queens County, New York, officers allegedly observed the defendant driving a black Infiniti, stop and illegally park in a bus stop, run into a store, get back into the car and continue driving. Shortly thereafter, the officers conducted a car stop, learned the defendant had a suspended license, recovered a switchblade from the defendant's person and arrested the defendant. The car was taken to the precinct, an inventory search was conducted, and a loaded Century International Arms 9-millimeter firearm was recovered from inside a bag in the trunk.

Thereafter, ECT swabbed the firearm for DNA evidence in three areas: (1) the entire grip area; (2) the trigger/trigger guard; and (3) the slide groove/release. The swabs were submitted to the New York City OCME, Department of Forensic Biology for DNA testing. A cup that the defendant used at the precinct was also sent to OCME for testing.

Later that same day, the defendant was arraigned in Criminal Court on a felony complaint. On May 22, 2023, the defendant was indicted and charged with Criminal Possession of a Weapon in the Second Degree, C.P.L. § 265.03 (1)(b); Criminal Possession of a Weapon in the Second Degree, C.P.L. § 265.03 (3); Criminal Possession of a Weapon in the Third Degree, C.P.L. § 265.02(8); Criminal Possession of a Firearm, C.P.L. § 265.01(B)(1); Criminal Possession of a Weapon in the Fourth Degree, C.P.L. § 265.01(1); Aggravated Unlicensed Operation of a Motor Vehicle in the Third Degree, V.T.L 511(1)(A); and Operating or Driving a Motor Vehicle without a License V.T.L. 509(1). The defendant was arraigned on the indictment in Supreme Court on June 14, 2023.

DNA testing was conducted on the swabs from the firearm. An OCME report, dated June 6, 2023, indicated that the grip area and the trigger/trigger guard areas "yielded a mixture, best described as having five or more contributors." Due to OCME laboratory policies, no interpretation or comparisons were made to the mixture from those swabs. The same report indicated that the swab of the "slide groove/release" "yielded a mixture that was estimated to have three contributors," although none of its "individual contributors could be determined," and that the mixture "was suitable for comparison."

On June 12th, 2023, an OCME report indicated that the lab had "developed a DNA profile from the cup attributed to the defendant" and a comparison of the cup profile to the slide groove/release mixture "led to a likelihood ratio of 1.32 million." On November 18, 2024, the Court ordered the defendant to submit to a buccal swab to obtain a reliable ('known') DNA sample for comparison. On January 7, 2025, OCME issued a laboratory report on the comparison of the defendant's profile to the swab of the "slide groove/release" which indicated that the likelihood ratio that the sample originated from the defendant and two (2) unknown persons was 1.15 million times more probable than if it originated from three (3) unknown persons. OCME therefore concluded that the defendant was included as a contributor to the sample.

Criminal Courts routinely grapple with the admissibility of DNA evidence and expert testimony. On trial cases involving recovered firearms, evidence is routinely offered regarding the recovery, processing and the testing of the firearm. Experts are called to testify about the steps that they take to secure and unload the firearm, the tests that are conducted to determine if [*3]the firearm is operable and if any DNA or fingerprints were identified on the firearm. NYPD witnesses, usually qualified as experts, testify as to their qualifications, NYPD firearm protocols, including checking for fingerprints, the improbability of finding fingerprints, and checking for DNA which they explain involves swabbing three specific areas of the firearm and sending the swabs to the OCME laboratory for testing.

Personnel/Analysts from OCME, often qualified as experts, testify about their qualifications, the lab's accreditation, the process used to make sure the lab and the samples are not contaminated and the process they employ to create and analyze a DNA profile or comparison. They give the trier of fact an overview of what DNA is, the markers that they use, and all the steps involved, including extraction, quantification, amplification, separation, analysis and interpretation. Experts also testify as to the reviews that are required and the accuracy percentage of the tests. Armed with the reports, they often conclude by giving an opinion on whether the defendant is included in the sample, the percentage of his or her DNA on the sample compared to anyone else and/or why a sample is or was not suitable for comparison.

OCME analysts often utilize and rely on software for comparisons. One essential software is STRmix, which is defined as "a fully continuous probabilistic genotyping forensic software which combines biological modeling with mathematical processes ^[FN1] ." The software 1) interprets and attempts to deconvolute DNA profiles, put more simply, uses a computational method to separate and identify individual genetic profiles or cell types from a mixed sample of contributors found on evidence samples, 2) conducts comparisons between a suspect's sample and evidence samples and provides statistical weights, known as 'likelihood ratios' and 3) calculates likelihood ratios for quality control purposes.

Defendant does not dispute the reliability and acceptance of the use of STRmix in general but instead questions the step before any data is processed by STRmix, which requires an analyst to estimate and input the number of contributors to a sample. He avers that the accuracy of the number input (the estimate), effects the accuracy of the tests. He asserts that it is common to misidentify and underestimate the number of contributors and the higher number of contributors, the less accurate the analysis. He highlights that his argument is supported by OCME, whose protocol prohibits testing samples determined to have five or more contributors ^[FN2] , due to the high [*4]error and inaccuracy rates.

Specifically, he challenges the analyst's conclusions here that the slide groove release had three contributors, arguing that the analyst could have underestimated the number of contributors, and if so, the sample could have five or more contributors, rendering it unsuitable for testing or at least, less reliable. In support of his under-estimation theory, he points out that the other two locations of the firearm that was swabbed, the grip area and the trigger/trigger guard, resulted in a determination that those locations had five or more contributors. He also avers that the error rate for a five-person mixture appearing as a three mixture by analysts were seventeen and a half percent.

In addition, for the Court's consideration, he provided 1) a 2021 OCME report that indicates that their analysts have "an extremely difficult time distinguishing complex DNA mixtures of three, four and five contributors," 2) the OCME internal validation of STRmix v 2.7 report, specifically experiment number 4, where "various hypotheses were tested using two-person, three-person, four person, and five-person mixture sets based on the determined apparent NOC" (number of contributors), which he alleges shows that analyst estimates were more likely to be incorrect when there were more contributors in a mixture, 3) OCME's error rates, indicating that it is incorrect roughly seventeen percent of the time for three person mixtures, forty-six percent of the time with four person mixtures and 87 percent of the time for five person mixtures, 4) a 2015 Coble study which he argues shows that eighty-six percent of six person mixtures present as having five or fewer contributors, and 5) a 2023 Hicklin study that "noted that most labs have validated four-person mixtures or less for STRmix."

Analysis

It is clear to this Court, both by a review of cases in which Courts have examined the use of STRmix and the scientific articles defendant submitted as exhibits, that the defendant is correct in his assertion that STRmix requires an analyst to estimate the number of contributors in a sample and that any inaccuracy with the estimate, impacts the accuracy of the STRmix results. Moreover, five or six person mixtures, have a high percentage of error margin, are unreliable and therefore are not tested by most laboratories. However, defendant's other contention, that the DNA sample here, which was estimated by the analyst to be a three person mixture, is likely an underestimation and thus as unreliable as a five or six person mixture, requiring exclusion or a Frye hearing, is not as clear.

As this Court understands it, determining the number of contributors of a mixture, requires several considerations, including the number of loci locations examined (specific physical locations of DNA sequences found along the DNA molecule), the number of potential contributors involved, the level of variables/differences present in each profile and the training and expertise of the analyst and/or anyone reviewing the examination. Accordingly, this step or the science in this area, is not perfect or error proof. However, the fact that there is an estimation and/or a potential for inaccuracy, does not negate the scientific acceptance or the reliability of this DNA evidence.

Turning to the studies submitted to the Court by the defendant in support of his motion. [*5]First, the 2015 Coble report ^[FN3] , authored by Michael D. Coble, Jo-Ann Bright, John S. Buckleton and James M. Curran. This research attempted to quantify in mathematical/algebraic terms the percentage of underestimating the number of contributors in a mixed DNA sample.

In the article, the authors explain that the crucial first steps of DNA analysis require determining if DNA is present and if so, the number of contributors in a sample/mixture. To determine the number of contributors to a mixture, analysts look at certain loci, specific physical locations of DNA sequences found along the DNA molecule. The FBI suggests thirteen specific such locations, that are considered 'highly polymorphic,' meaning that they show a high level of variation, which is the accepted law enforcement standard. By looking at these locations, an analyst is supposed to be able to create a unique profile for each contributor and thus identify and differentiate between each contributor.

Other studies have examined the benefit of expanding the number of locations that are examined. The authors however accept that different methods of interpretation use different amounts of profile information. They also explain that likelihood ratios require the assignment of the number of contributors to a profile, which is why they sought to examine the accuracy of the assignment (estimation). In trying to determine the accuracy of this assignment, the authors considered what they referred to as the four major race groups in the US population, African American, Asian, Caucasian and Hispanic, and how that factored into the accuracy of the quantification of the number of contributors and/or inaccuracy of the estimate.

They concluded that 1) the true number of contributors on a mixture that appears to have five or six estimated contributors is unlikely to be one or two, 2) that the true number of contributors on a mixture with a higher number of contributors is highly likely to appear as less contributors, 3) a six person profile has a 0.9999 percent probability of being estimated as a five person or less person profile when using 13 loci locations, and a 0.8599 percent probability when using more loci locations, and their ultimate conclusion, 4) when there is a five or six person mixture, it is beneficial to examine a higher number of loci locations.

Second, the "Variation in assessments of suitability and number of contributors for DNA mixtures ^[FN4] ," referred to by the defendant as the Hicklin study, was authored by R. Austin Hicklin, Nicole Richetelli, Brandi L. Emerick, Robert A. Bever and Jonathan M. Davoren in 2023. In the study, 29 DNA mixtures were sent to 134 different analysts at 67 forensic laboratories to be analyzed for suitability for comparison/analysis and to determine the number of contributors to the sample. The authors noted that the lab policies and procedures to determine both, varied. They indicated that their research was conducted in part "to provide key empirical data to the legal and forensic science communities" and to address concerns regarding complex DNA mixtures, considering the advancements in probabilistic genotyping and "the widespread adoption in forensic laboratories" since the last studies/reports.

In their paper, they refer to DNA analysis as being the "gold standard" in forensic science, but clarify that refers to DNA samples that are single source samples and/or simple [*6]mixtures, which they indicate include two person mixtures/contributors, samples with a relatively high DNA quantity and samples with no degradation (break down into smaller fragments) or inhibition (issues with replication of DNA). They point out that "several reports and reviews have called for additional research into the reliability of the interpretation of the "complex mixture samples" which they define as including three or more person mixtures/contributors, have low amounts of DNA quantity, exhibit degradation or inhibition and/or contain contributions from individuals with shared alleles.

They begin their analysis by asserting that DNA laboratories receive samples and do not always know if it is a single source or a mixture or if the mixture is simple or complex. Therefore, they point out, that an analyst must evaluate the sample to determine suitability for analysis and estimate the number of contributors to the sample. They also indicated that "[t]hese initial assessments impact whether a DNA mixture is analyzed at all, and if so, how the interpretation, comparison, and statistical analysis proceeds given the profile characteristics ^[FN5] ."

They concluded that on the question of suitability for comparison/analysis, the results varied depending on the particular laboratory's policy and procedures (referred to by them Standard Operating Procedures), and even amongst laboratories with the same policy and procedures, there was only a 66% agreement rate. They also concluded that for labs following the same policy and procedures, the number of contributors estimated was 79% correct, which dropped to 63% when different laboratories were involved. They also concluded that both labs had a 7% incorrect rate. Despite these results, they concluded, "[i]ncorrect NoC (number of contributors) assessments have been shown to affect statistical analyses in some cases, but do not necessarily imply inaccurate interpretations or conclusions" and "[m]ost incorrect NoC estimates were overestimates, which previous research has shown have less of an effect on likelihood ratios than underestimates.^[FN6] "

Third, OCME experiment number 4 seems to indicate that "determining the NOC is complicated by allele sharing and various artifacts that may be present from amplification [copying] and/or capillary electrophoresis [separation]^[FN7] ." They also note that "there are various reasons as to why the true and apparent NOC (number of contributors) may not align, leading to an under- or over-estimation of the true NOC ^[FN8] ." The report explains that such things as increased number of artifacts or elevated stutter peaks can present as additional contributors, while low template amounts, degradation, trace contributors with drop-out or potential allele sharing, can lead to a determination that it was fewer contributors. If the Court understands the terminology correctly, the OCME experiment points out several factors that may impact on the accuracy of the estimate, at each step of the analyst's analysis of the number of contributors in a mix, and the specific factors that can lead to an overestimation or underestimation.

In short, as pointed out by the defendant and these studies, numerous variables may [*7]impact the accuracy of the STRmix results and there is a direct correlation between the accuracy of the analyst's estimation of the number of contributors and the accuracy and the reliability of the results. It is also possible that a sample that an analyst estimates as a three person mixture may have additional contributors, which would affect the error rate for the test and if the number was underestimated by two or more, would affect whether the comparison could even be made.

The guidelines governing the admissibility of expert testimony in New York based on the use of new or novel scientific theories or techniques is enumerated by the rules set forth in Frye v. United States, 293 F. 213 (D.C. Cir. 1923). In determining whether scientific evidence is admissible under Frye, a court must assess the "... elemental question of whether the [challenged] techniques, when properly performed and generate results accepted as reliable within the scientific community." People v Wesley, 83 NY2d 417, 422 (1994). See People v. Wakefield, 38 NY3d 367 (2022). The relevant scientific community consists of individuals who have knowledge of the particular scientific or technical area involved in developing, studying and understanding the techniques. See, Wesley, supra. The particular procedure need not be unanimously endorsed by the scientific community but must be generally accepted as reliable. See People v. Middleton, 54 NY2d 42 (1981).

A Frye hearing is necessary only if the expert testimony involves novel or experimental matters. See, People v. Byrd, 51 AD3d 267 (1st Dept. 2008). Whether a particular technique represents a new or novel science is a separate and distinct inquiry from whether the scientific technique was applied correctly in a specific case. See, Parker v. Mobil Oil Corp., 7 NY3d 434 (2006). The Frye test concerns only the acceptability and reliability of the scientific technique; not the adequacy of specific procedures used to generate the particular evidence sought to be admitted. See, Wesley, supra. Moreover, the modification of an existing test does not constitute a new or novel science. See, Parker, supra; Styles v. General Motors Corp., 20 AD3d 338 (1st Dept. 2005). Where a particular technique is found to be generally acceptable by the relevant experts, a Frye hearing is usually not warranted.

Expert testimony on the science of DNA has been routinely admissible in the New York courts. See, Wesley, supra. Indeed, DNA experts routinely testify about statistical probabilities at trial. See, People v. Brown, 13 NY3d 332 (2009); see, e.g., People v. Megnath, 27 Misc 3d 405 (Sup. Ct., Queens Cty. 2010). Moreover, as pointed out by the People, the STRmix program had been the subject of numerous peer-reviewed journal articles and has been evaluated, approved and adopted by fifty-seven (57) laboratories in the United States, including the Federal Bureau of Investigation ("FBI"), all nine (9) State and Territory laboratories in Australia, and thirteen (13) labs internationally. Furthermore, in People v. Bullard-Daniel, the Court found that the People established that the methods employed in the STRmix program were generally accepted as reliable within the relevant scientific community. See People v. Bullard-Daniel, 203 AD3d 1630 [4th Dept 2022].

In addition, other courts have conducted Frye and Daubert ^[FN9] hearings on the use of STRmix and found it to be the reliable and accepted by the scientific community. In People v Bullard-Daniel, the Court held a three-day hearing where it considered "voluminous submissions" including "cases, articles from scientific journals, and expert affidavits" along with expert testimony, the findings of several scientific organizations and the acceptance by the DNA [*8]Subcommittee, which is "made up of experts in various scientific disciplines related to DNA analysis." See People v Bullard-Daniel, 203 AD3d 1630 (4th Dept 2022). See also People v Middleton, 54 NY2d 42, 49 [1981]) and People v. Muhammad [14th Cir.Ct., Muskegon Co., Dec. 15, 2015] and also upheld the admissibility of the DNA test results.

Here, the defendant cites to People V. Williams, 35 NY3d 24, (2020), where the Court of Appeals held that that the trial Court had abused its discretion in allowing DNA testimony without conducting a Frye hearing. The DNA issue there was that the DNA sample was insufficient for standard testing and required increasing the amount of amplification/copying cycles from 28 to 31, known as LCN typing. An expert indicated to that Court that LCN typing was not generally accepted or reliable, presented several articles that concluded likewise and indicated that only one publicly funded laboratory conducted such testing. Therefore, the Williams decision is distinguishable from the instant case, since numerous courts have conducted comprehensive Frye hearings on STRmix, considered oral arguments and written submissions and concluded that STRmix is reliable and generally accepted in the scientific community. Moreover, STRmix is extensively used and relied upon by accredited laboratories nationwide.

Defendant's concerns as to the accuracy of the analyst estimates of the number of the contributors, the percentage of error of tests involving multiple contributors and the greater degree of inaccuracy/error with the higher number of contributors, are no different than in any other case where STRmix and expert testimony has been admitted by Courts across New York State. Here, defendant has failed to raise any issues that specifically effected the accuracy of the tests in this matter, such as the analysts training, the process followed, any disagreements between the analyst and the reviewer, issues of multiple replications or similarity of contributors, anything that would warrant a departure from the general acceptance and admissibility of this evidence and testimony.

Moreover, this Court agrees with the People, and concludes that the issues defendant raises, do not warrant preclusion of the evidence or of the testimony. The issues including the studies and error rates can be addressed in cross examination and/or by testimony of a defense expert.

Based upon the foregoing, the defendant's motion to preclude DNA evidence or, in the alternative, for a Frye hearing, is DENIED.

As to the defendant's request for special jury instructions, that is RESERVED FOR THE TRIAL COURT

This constitutes the decision of the Court.